The prevalence of polyps of the large intestine in Oslo: an autopsy study.
ABSTRACT The colon and rectum were removed from 445 consecutive autopsies and examined for the presence of polypoid lesions under an illuminating magnifying lens. There were seven clinically undiagnosed carcinomas, 329 adenomas, 174 hyperplastic polyps, 59 mucosal tags, 34 other polypoid lesions, and 54 polyps in which no histologic diagnosis could be made. Adenomas were present in 34% of the men and 32% of the women. The adenoma prevalence rate increased smoothly with age in men, whereas the rate in women showed a biphasic variation with a peak at 50--59 years followed by a second increase after 70 years of age. The distribution of adenomas within the bowel showed a marked change with age, from a moderate left-sided predominance before 60 years of age to a marked right-sided predominance in patients over 80 years of age, particularly in men. Hyperplastic polyps were present in 27% of the men and 18% of the women. In men, their frequency and distribution within the bowel showed similar changes with age as were seen for the adenomas. In women, no such age-dependency was observed for hyperplastic polyps. Multiplicity of polyps increased with age and was more pronounced in men than in women. Adenomas in women tended to be larger, more atypical, and more villous than in men. A comparison with a similar study from Northern Norway showed no difference in the prevalence of adenomas in spite of a 70% higher incidence rate for colorectal cancer in Oslo.
- [Show abstract] [Hide abstract]
ABSTRACT: OBJECTIVE:Computed tomography (CT) or magnetic resonance (MR) colonography is a new technique that uses data generated from CT or MR imaging to create two- and three-dimensional scans of the colon. It has been advocated to become the new primary technique of screening for colorectal cancer. The economic feasibility of such recommendation, however, has not yet been evaluated.METHODS:The cost-effectiveness of two screening strategies using CT colonography or conventional colonoscopy was compared by computer models based on a Markov process. We supposed that a hypothetical population of 100,000 subjects aged 50 yr undergoes a screening procedure every 10 yr. Suspicious findings of CT colonography are worked-up by colonoscopy. After polypectomy, colonoscopy is repeated every 3 yr until no adenomatous polyps are found.RESULTS:Under baseline conditions, screening by CT colonography costs $24,586 per life-year saved, compared with $20,930 spent on colonoscopy screening. The incremental cost-effectiveness ratios comparing CT colonography to no screening and colonoscopy to CT colonography were $11,484 and $10,408, respectively. Screening by colonoscopy remains more cost-effective even if the sensitivity and specificity of CT colonography both rise to 100%. For the two screening procedures to become similarly cost-effective, CT colonoscopy needs to be associated with an initial compliance rate 15–20% better or procedural costs 54% less than colonoscopy.CONCLUSIONS:To become cost-effective and be able to compete with colonoscopy in screening for colorectal cancer, CT or MR colonography would need be offered at a very low price or result in compliance rates much better than those associated with colonoscopy.The American Journal of Gastroenterology 01/1999; 94(8):2268-2274. · 9.21 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: Aim: Polyp is called formations protruding into the lumen in the gastrointestinal tract. In this study, we aimed to determine frequency, anatomic distribution within the large bowel and pathologic types of colon polyps based on the colonoscopy results. Materials and methods: The colonoscopy reports between 2010 and 2011 were analyzed retrospectively. Complaint of the patients on admission, the presence of polyps, anatomic distribution of polyps, polyp type and polyp size were evaluated. Result: A total of 4000 colonoscopy reports were examined. The largest proportion of polyps occurred in the left colon (74.4%), followed by the ascending colon (11.5%). Eighty-five were adenomatous polyps, 71 were hyperplastic polyps, 12 were inflammatory polyps, 12 were serrated adenoma, 3 were adenocarcinoma, and 8 were tubulovillous adenomas. Discussion: In light of like this study, the development of screening and follow-up programs in Turkey will be useful to reduce the incidence and mortality of colorectal cancerEuroasian Journal of Hepato-Gastroenterology,. 01/2014;
- [Show abstract] [Hide abstract]
ABSTRACT: Our systematic review summarizes the evidence concerning the accuracy of serum diagnostic and prognostic tests for colorectal cancer (CRC).PLoS ONE 08/2014; 9(8):e103910. · 3.53 Impact Factor
The Prevalence of Polyps of the Large Intestine
An Autopsy Study
MORTEN H. VATN, MD' AND HELGE STALSBERG, MDt
The colon and rectum were removed from 445 consecutive autopsies and examined for the presence of
polypoid lesions under an illuminating magnifying lens. There were seven clinically undiagnosed car-
cinomas, 329 adenomas, 174 hyperplastic polyps, 59 mucosal tags, 34 other polypoid lesions, and 54
polyps in which no histologic diagnosis could be made. Adenomas were present in 34% of the men and
32% of the women. The adenoma prevalence rate increased smoothly with age in men, whereas the rate
in women showed a biphasic variation with a peak at 50-59 years followed by a second increase after
70 years of age. The distribution of adenomas within the bowel showed a marked change with age,
from a moderate left-sided predominance before 60 years of age to a marked right-sided predominance
in patients over 80 years of age, particularly in men. Hyperplastic polyps were present in 27% of the
men and 18% of the women. In men, their frequency and distribution within the bowel showed similar
changes with age as were seen for the adenomas. In women, no such age-dependency was observed for
hyperplastic polyps. Multiplicity of polyps increased with age and was more pronounced in men than
in women. Adenomas in women tended to be larger, more atypical, and more villous than in men. A
comparison with a similar study from Northern Norway showed no difference in the prevalence of
adenomas in spite of a 70% higher incidence rate for colorectal cancer in Oslo.
Cancer 49:819-825. 1982.
P of adenomas as precursors of carcinomas of the
large intestine,'-' but many aspects of the polyp-cancer
sequence need further study. The incidence of adenomas
and carcinomas of the colon is high in western Europe
and the United state^,^ in contrast to the low incidence
of both among populations of Afro-Asian
Within Norway, the incidence of colorectal cancer is
about 70% higher in Oslo than in Northern Norway.'
In the present investigation we have studied the prev-
alence and distribution of colorectal polyps in patients
coming to autopsy at Ulleval Hospital, the largest gen-
eral hospital in Oslo. A similar study from Northern
Norway has been published previo~sly.~
REVIOUS studies have emphasized the significance
From the Department of Pathology, Ulleval Hospital, University
of Oslo, and Institute of Medical Biology, University of Tromss.
* Former Resident in Pathology.
t Professor of Morphology.
Address for reprints: Dr. Morten H. Vatn, Rikshospitalet, Oslo 1,
The authors thank Mrs. Vivi Jaregg, Miss Grethe Ssrensen, and
Mr. Arne Nilsen for technical assistance, Mr. Ingar Holme for sta-
tistical advice, and Dr. Tor J. Eide for permission to use unpublished
data on distribution of adenomas by size in Northern Norway in
comparisons with present data.
Accepted for publication January 7, 198 1.
Materials and Methods
During the period from 1972-1973 the entire large
bowel was removed from 100 consecutive autopsies in
each of the age groups 50-59, 60-69, 70-79, and 80
years and older. During the same period, 45 large in-
testines from autopsies on patients under 50 years were
collected. The total material is from 181 women and
264 men. Because of the relative rarity of lesions and
small number of patients under 50 years of age, the
analyses are mostly concerned with the 400 patients 50
years and older. The intestines were opened, washed
with tap water, and examined unfixed under an illu-
minating magnifying lens (X2). The total length of the
large bowel, the distance from the cecal pole to each
polypoid lesion, and the largest diameter of each lesion
were measured. In addition, the length of each anatom-
ical segment of the bowel was measured in ten individ-
uals from each age group.
All polypoid lesions were removed, packed in filter
paper or gauze, fixed in 4% formaline and embedded
in paraffin. One central section through each lesion was
stained with hematoxylin and eosin. Histologic classi-
fication was done by both authors according to the
WHO classification," supported by the criteria of Lane
19 $0.85 0
American Cancer Society
CANCER February 15 1982
I. Number and Types of Polypoid Lesions of the Large
Intestine in 445 Consecutive Autopsies
M ucosal tag
Other polypoid lesions
No histologic diagnosis
2. Multiplicity of Polypoid Lesions at Gross Examination,
by Sex and Age
No. of polyps per intestine
examined Sex Age
et a/." for the distinction of tubular adenomas from
hyperplastic polyps. A histologic diagnosis was given
only when both observers had independently found the
material adequate for diagnosis.
For each sex and age combination, X2-tests with
Yates' correction were used to test the hypothesis of an
aggregation of adenomas in proximal and distal parts
of the bowel. The tests were done separately for solitary
and multiple adenomas. The number of adenomas
within the proximal 30% and within the distal 70% of
the length of the large intestine were counted and tested
3. Prevalence of Adenomas by Age and Sex
adenoma Age Rate
against the hypothesis of an 30-70% distribution. An
analogous test procedure was done for the distal 30%
versus the proximal 70% of the bowel length.
A total of 657 polypoid lesions were identified in 218
of the 445 individuals (Table 1). Of these, 50% were
adenomas, 26% hyperplastic polyps, and 1% carcino-
mas. Mucosal tags with normal mucosal histology made
up 9%, and in 8% of the gross lesions no histologic
diagnosis could be made, mostly because of advanced
autolysis. Among the adenomas, 9% were classified as
tubulovillous and 0.6% as pure villous adenoma. The
seven carcinomas were all clinically undiagnosed. They
occurred in six patients, of whom four were 80 years
of age or older. The mean size of the carcinomas was
4.8 cm in the latter and 1.6 cm in the two younger
patients. The largest two carcinomas, 6 cm each, were
in the ascending colon; two of the smallest tumors, 2.1
and 1.2 cm, respectively, were in the rectum; and the
remaining three were in the sigmoid. Other lesions in-
cluded lipomas, fibromas, leiomyomas, metastases, in-
flammatory and vascular lesions. Of all polypoid lesions,
67% were sessile. Among intestines with polypoid le-
sions, 87 (40%) had only one single lesions, and the
highest number of polypoid lesions in one specimen was
16. There is a moderate trend of increasing multiplicity
with age (Table 2).
Histologically verified adenomas were present in 9 1
men and 57 women (Table 3). The prevalence rate in
men increases steadily up to 70 years of age, and then
remains at the same level in the oldest age group.
The rate also increased with age in women, except
for a relatively high value in the 50-59 year group. The
female rate is higher than in men under 60 years of age
and lower than in men after that age. Only in the com-
bined 60-79 age group is the sex difference statistically
significant at the 5% level. The trends are similar, but
a little more accentuated, when the mean number of
adenomas per intestine examined is considered, and also
when only adenomas 5 mm and over or 10 mm and over
are considered (Table 4).
The distribution of adenomas within the large intes-
tine is shown in Figure 1, as related to 10% lengths of
the distance from the cecal pole to the pectinated line.
The average relationship between these percentual
lengths and the anatomic segments are shown at the
bottom of the figure. Before 60 years of age, there are
more adenomas in the distal than in the proximal half
of the bowel. After 70 years of age, in both sexes, and
particularly in men over 80, there is a marked prepon-
derance of right-sided adenomas. Thus, in the proximal
30% of the bowel, including the cecum, the ascending
and the first part of the transverse colon, there is a
POLYPS OF LARGE INTESTINE
Vatn et al.
marked increase in the frequency of adenomas with age
in men, and a similar, but smaller increase in women.
In the distal 30% of the bowel, including most of the
sigmoid and the rectum, there is no clear age trend in
the frequency of adenomas in women. In men, there is
a moderate increase up to the 70-79 year age group,
and then a remarkable decrease in the oldest age group,
possibly indicating a regression of distal adenomas in
With the chi-square test as described in the methods
section, there is a significant ( P < 0.05) aggregation of
solitary adenomas in the distal 30% of the bowel in men
and for both sexes combined in the 50-59 and 60-69
year age groups.
The differences for women and for multiple adeno-
mas are in the same direction in these age groups, but
do not reach the 5% level of significance. In the proximal
30%, the aggregation of adenomas was statistically sig-
nificant in both sexes only at 80 years and older, and
only for multiple adenomas.
Of all adenomas, 36% were pedunculated. No ade-
noma over 9 mm was found in the distal 30% of the
bowel. Otherwise, the distribution of small and large
adenomas were similar to that of all adenomas (Fig. 2).
The adenomas in women tended to be larger, more
atypical, and more villous than in men. Thus, the pro-
portion of adenomas larger than 9 mm was 21% for
women and 8% for men 50-69 years of age, and 21%
for women and 15% for men 70 years and older. Severe
atypia was diagnosed in 13% of the adenomas in women
and 4% of the adenomas in men. The proportion of
adenomas classified as villous or tubulovillous was 24%
in women and 12% in men 50-69 years of age, and 9%
in women and 7% in men 70 years and older. The an-
atomic distribution of villous and tubulovillous adeno-
mas was similar to that of all adenomas.
The frequency of hyperplastic polyps shows a mod-
erate increase with age in men, but appears to be in-
dependent of age in women (Table 5). This applies both
to the rate of intestines carrying one or more hyper-
plastic polyps, and to the mean number of hyperplastic
polyps per intestine examined. Both frequencies are
generally higher in men than in women. The anatomic
distribution of hyperplastic polyps in men has a similar
age trend, as was seen for adenomas, with a shift from
a predominance of distal lesions in the younger groups
to a predominance of proximal lesions in those over 80
years of age (Fig. 3). In women, hyperplastic polyps are
more evenly distributed throughout the large bowel at
Nonepithelial polyps, consisting of 17 lipomas and
17 others (fibromas, leiomyomas, vascular, and inflam-
matory polyps), showed the same distribution as ade-
nomas and hyperplastic polyps, with a concentration at
both ends of the large intestine.
4. Mean Number of Adenomas per Ten Intestines
Examined, by Age, Sex, and Adenoma Size
and over Age All sizes
8 0 f
FIG. I. Mean number of adenomas per 10 intestines examined, per
10% lengths of the large intestine from the cecal pole to the pectinated
line, in different age groups in males (left) and females (right). The
average extension of each anatomical segment of the intestine is in-
dicated on the bottom line. CA-cecum and ascending; T-transverse;
D-descending; S-sigmoid; R - rectum.
CANCER February 1.5 1982
The common opinion that most polyps arise in the
distal part of the large intestine has been supported by
several studies,'*-'' whereas others have found that most
polyps arise in the proximal part of the
present material, both adenomatous and hyperplastic
polyps were more often found in the sigmoid and rectum
in the younger age groups, whereas both types more
often arose in the proximal part in age groups 70 years
of age and older. This is an agreement with the previous
study from Northern Norway.' The changing distri-
bution of polyps with age corresponds well with the
TABLE 5. Prevalence of Hyperplastic Polyps (HP) and Mean
Number of HP per Ten Intestines Examined, by Age and Sex
Mean no. of
HP per 10
with HP Age Rate
FIG. 2. Number of adenomas of
the large intestine in 400 patients
50 years and older, per 10% lengths
of the large intestine, by size.
distribution of carcinomas. For most age groups, most
carcinomas occur in the rectosigmoid region. In the very
old, however, there is a particular increase in carcino-
mas of the proximal c~lon.''-~~
strengthen the theory of a causal relationship between
carcinoma and both adenoma and hyperplastic polyps,
although the relationship may be different for the two
types of polyps.
Our observations support the conclusion of Potet and
So~llard'~ that adenomas are fewer but more malignant
in women than in men. Female adenomas showed more
marked atypia, more villous growth, and larger size in
the present material, all of which are associated with
increased malignant potentiaL2'
The present study is largely comparable to that al-
ready published from Troms County in Northern Nor-
way,' although some minor differences in methodology
were introduced when the latter study was linked to an
international cooperative project. According to data
from the Cancer Registry of Norway, the age-adjusted
incidence rate for cancer of the colon is 38.6 per 100,000
for men and 30.7 for women in Oslo, and 16.8 for men
and 21.2 for women in Trams.* For cancer of the rec-
tum, the rates are 21.1 for men and 13.0 for women in
Oslo, and 13.4 for men and 8.9 for women in Troms.
If the prevalence of adenomas were a predominating
indicator of colorectal cancer risk in the population, one
would expect to find substantial differences between the
Figure 4 shows that the prevalence of intestines with
adenoma is almost identical in Oslo and Troms. Inter-
POLYPS OF LARGE INTESTINE
. Vatn el al.
estingly, both show a similar difference between male
and female rates, with female rates being higher than
the male rates before ages 55-60, and lower than male
rates after that age. This sex difference is reminiscent
of that for the incidence rates for cancer of the colon,
as shown for Norway as a whole in Figure 5. The female
and male curves for adenoma prevalence rates intersect
at 55 (Troms) and 58 (Oslo) years of age, whereas the
colon cancer incidence rates intersect at 64 years. The
difference of 6-9 years may well be related to the av-
erage lead time from the presence of adenoma to the
diagnosis of carcinoma in the patients in whom can-
cerous development occurs.
The main difference in methodology between the
Oslo and Troms studies is that the intestine was ex-
amined fresh in the present study, and after fixation in
the Troms study, whereas the same type of illuminating
magnifying lens was used in both. Figures 6 and 7 show
the mean number of adenomas per 10 intestines ex-
amined by sex and age, for all sizes and for adenomas
5 mm in diameter and larger only. It seems unlikely
that adenomas 5 mm and larger have been missed to
any significant extent in any of the studies. As fixation
makes the tissues shrink slightly, its effect on the num-
ber of adenomas over a given size would be a reduction
rather than an increase. Nevertheless, there is no in-
dication that adenomas 5 mm and over are more fre-
quent in Oslo than in Troms. For adenomas 10 mm and
over, the rates for all ages and both sexes combined
were 1.0 per 10 intestines in Oslo and 0.9 per 10 in-
testines in Troms. In conclusion, there is no evidence
that adenomas of the large bowel has a higher preva-
lence in Oslo than in Troms, in spite of a 70% higher
colorectal cancer incidence rate.
Discrepancies between adenoma prevalence rates and
cancer incidence rates are also seen in a comparison of
men and women, and in a comparison of the colon and
the rectum. If the points of intersection between male
and female rates are taken as reference points, the
male:female ratios 10 and 15 years after this point are
2.1 and 1.7 for adenoma prevalence rates and only 1.1
and 1.2 for colon cancer incidence rates. The corre-
sponding sex ratios for the mean number of adenomas
per intestine examined are even greater (4.1 and 3.2 for
adenomas of all sizes; 3.7 and 3.7 for adenomas 5 mm
A comparison of colon and rectum shows that rectal
cancers constitute 32% of all colorectal cancers in Oslo
and 36% in Trams.* In contrast, adenomas in the distal
tenth of the large intestine, which includes the rectum
and the distal few centimetres of the sigmoid, constitute
only 9.6% and 8.0% of all colorectal adenomas in Oslo
FIG. 3. Mean number of hyperplastic polyps per 10% lengths of the
large intestine. in different age groups in males (left) and females
(right). Bottom line: CA-cecum and ascending; T--transverse; D-de-
scending; S-sigmoid; R-rectum.
/ - : i Y o
I , .
60 70 80
FIG. 4. Adenoma prevalence rates in Oslo (current study) and
Tromse, Northern N ~ r w a y , ~ by sex and age.
CANCER February 15 1982
FIG. 5. Annual incidence rates for cancer of the colon and rectum
in Norway, by sex and age. Data from The Cancer Registry of Nor-
TromrB. al I
FIG. 6. Mean number of adenomas per 10 intestines examined in
Oslo and Tromse,’ by age (males). Separate curves for adenomas of
all sizes and adenomas 5 mm and over only.
and Troms, respectively. If only the younger age groups
are considered, the percentage of adenomas in the distal
tenth of the large intestine is still not more than 16%
in Oslo (under 60 years) and 14% in Troms (under 65
years). On the assumption that all or the great majority
of colorectal cancers develop from adenomas, it appears
therefore that the rate and risk of malignant develop-
ment per adenoma is greater in Oslo than in Troms,
greater in women than in men, and greater in the rec-
tum than in the colon. These observations fit with a
two- or multistep development of carcinoma, in which
the development of adenoma is only one step, and some
of the observed differences in cancer incidence may be
caused by differences in the factors promoting further
development from adenoma to carcinoma.
A greater risk of malignant development per ade-
noma in Oslo than in Troms is also supported by the
finding of carcinomas in six of 445 patients ( 1.3%) in
the Oslo material and in only two of 280 patients (0.7%)
in the Troms material. Tubulovillous and villous pat-
terns were also diagnosed more frequently among the
adenomas in Oslo ( 10%) than in the Troms (1.4%).
However, although one of the authors participated in
typing of both materials, the studies were not done at
the same time, and the possibility of a shift in diagnostic
criteria on this point cannot be excluded. The distri-
bution of 31 tubulovillous adenomas in the Oslo ma-
FIG. 7. Mean number of adenomas per 10 intestines examined in
Oslo and Tromse,’ by age (females). Separate curves for adenomas
of all sizes and adenomas 5 mm and over only.
POLYPS OF LARGE INTESTINE . Vatn et al.
terial showed the same distribution as tubular adenomas
in the whole material.
A relatively high adenoma prevalence rate in the
younger group of women followed by a lower rate
around the 60-65 age group is seen in both Norwegian
materials, and was also observed in the study of Stem-
mermann and Yatani from Hawaii.”
Because this phenomenon is only seen in women, it
may possibly be related to the hormonal changes at the
age of menopause, which is particularly interesting with
regard to the strong epidemiologic association between
incidence rates for carcinomas of the colon and the fe-
1. Correa P, Duque E, Cuello C, Haenszel W. Polyps of the colon
and rectum in Cali, Columbia. Int J Cancer 1972; 9236-96.
2 . Morson B. The polyp-cancer sequence in the large bowel. Proc
Roy Soc Med 1974; 67:451-457.
3. Kozuka S, Nogaki M, Ozeki T, Masumori S. Premalignancy of
the mucosal polyp in the large intestine: 11. Estimation of the periods
required for malignant transformation of mucosal polyps. Dis Col
Rect 1975; 18(6):494-500.
4. Ekelund G, Lindstrem C. Histopathological analysis of benign
polyps in patients with carcinoma of the colon and rectum. Gut 1974;
5. Berg JW, Downing A, Lukes RJ. Prevalence of undiagnosed
cancer of the large bowel found at autopsy in different races. Cancer
6. Bremner CG, Ackerman LV. Polyps and carcinoma of the large
bowel in the South African Bantu. Cancer 1970 26:991-999.
7. Eichhorn RD, Bentlif PS, Kelsey JR. Ethnic differences in oc-
currence of rectal polyposis in over 6000 proctoscopies. Digestive Dis
8. Cancer Registry of Norway. The incidence of cancer in Norway
1972-76. Oslo, The Norwegian Cancer Society, 1978.
9. Eide TJ, Stalsberg H. Polyps of the large intestine in northern
Norway. Cancer 1978; 42:2839-2848.
10. Morson BC, Sobin LH. Histological typing of intestinal tu-
mours, international histological classification of turnours, no 15. Ge-
neva, World Health Organization, 1976.
11. Lane N, Kaplan H, Pascal RR. Minute adenomatous and hy-
perplastic polyps of the colon: divergent patterns of epithelial growth
with specific associated mesenchymal changes. Gastroenterology
12. Welch CE, McKittrick JB, Behringer G. Polyps of the rectum
and the colon and their relation to cancer. N Engl J Med 1952;
13. Mayo CW, Pagtaleman RJG. Malignancy of colon and rectum
in patients under 30 years of age. Surgery 1963; 53:711-718.
14. Martin JD Jr, Sutton G. Present status of polyps of the colon:
classification and malignant potential. Amer Surg 1965; 31:551-558.
15. Potet F. Soullard J. Polyps of the rectum and colon. Gut 1971;
16. Feyrter F. Zur Geschwulstlehre (nach Untersuchungen am
menschlichen Darm): Polypen und Krebs. Beitr Path Anat 1931;
17. Chapman I. Adenomatous polypi of large intestine: incidence
and distribution. Ann Surg 1963; 157:223-226.
18. Arminski TC, Mc Lean RW. Incidence and distribution of
adenomatous polyps of the colon and rectum based on 1.000 autopsy
examinations. Dis Col Rect 1964; 7:249-261.
19. Haenszel W, Correa P. Cancer of the colon and rectum and
adenomatous polyps. Cancer 197 1; 28: 14-24.
20. de Jong UW, Day NE, Muir CS, Barclay THC, Bras G, Foster
FH, et a/. The distribution of cancer within the large bowel. Int J
Cancer 1972; 10:463-477.
21. Morson BC, Bussey HJR. Predisposing causes of intestinal
cancer. Curr Probl Surg 1970; Feb:l-46.
22. Stemmermann GN, Yatani R. Diverticulosis and polyps of the
large intestine. A necropsy study of Hawaii Japanese. Cancer 1973;
3 I :1260-1270.
23. Howell MA. The association between colorectal cancer and
breast cancer. J Chron Dis 1976; 29:243-261.